Sheena A Josselyn’s research while affiliated with SickKids and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (191)


Astrocytes: emerging stars of engrams
  • Article

December 2024

·

7 Reads

Cell Research

·

Sheena A Josselyn

During learning experience, a subset of astrocytes in the hippocampus is activated and becomes both necessary and sufficient for subsequent memory recall. Williamson et al. discovered that these learning-associated astrocytes play an important role in memory by modulating engram neurons’ synaptic activity and facilitating the reactivation of engram neuron ensembles during memory retrieval.


Tagging and Manipulation of Fear Engrams

November 2024

·

8 Reads

The ability to manipulate neurons involved in memory formation and retrieval is a key factor to understanding neuronal circuits involved in cognition. In this chapter, we provide methods to tag and/or allocate neurons to a fear conditioning engram in the hippocampus and manipulate these neurons using optogenetic and chemogenetic approaches. For tagging, we describe IEG-based methods (TRAP2 mice or the viral RAM system) to express transgenes, such as designer receptors exclusively activated by designer drugs (DREADDs) or channelrhodopsin in fear engram cells. For allocation, the excitability of neurons in the CA1 region may be transiently and reversibly heightened using opsins, and heightened excitability increases the chance of the neurons being integrated into the fear conditioning engram. Allocated/tagged neurons then can be subsequently activated or silenced during the memory test using optogenetic or chemogenetic transgenes leading to heightened or decreased freezing, respectively.



FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5

+3

A sensitive period for the development of episodic-like memory in mice
  • Preprint
  • File available

November 2024

·

60 Reads

·

2 Citations

Episodic-like memory is a later-developing cognitive function supported by the hippocampus. In mice, the formation of extracellular perineuronal nets in subfield CA1 of the dorsal hippocampus controls the emergence of episodic-like memory during the fourth postnatal week (Ramsaran et al., 2023). Whether the timing of episodic like memory onset is hard-wired, or flexibly set by early life experiences during a critical or sensitive period for hippocampal maturation, is unknown. Here, we show that the trajectories for episodic-like memory development vary for mice given different sets of experiences spanning the second and third postnatal weeks. Specifically, episodic like memory precision developed later in mice that experienced early-life adversity, while it developed earlier in mice that experienced early-life enrichment. Moreover, we demonstrate that early-life experiences set the timing of episodic-like memory development by modulating the pace of perineuronal net formation in dorsal CA1. These results indicate that the hippocampus undergoes a sensitive period during which early-life experiences determine the timing for episodic-like memory development.

Download

On role models and Joe LeDoux

October 2024

·

4 Reads

Cerebral Cortex

Joseph LeDoux is a pioneering neuroscientist who has made profound discoveries that continue to impact our understanding of the neural basis of emotion and memory, particularly the role of the amygdala in threat conditioning. LeDoux’s trailblazing and elegant studies were some of the first to examine the circuit basis of behavior. His work combined techniques to trace pathways into and out of the amygdala important for threat conditioning and related behaviors. Since that time, these types of circuit tracing studies have exploded in popularity across neuroscience, and I would argue, we all owe a debt to LeDoux for this. LeDoux has made numerous additional contributions to neuroscience and, by bringing emotion back to neuroscience, has helped unite neuroscience with psychology. A gifted writer with a knack for communicating complicated scientific ideas in an accessible manner, LeDoux has become an ambassador of science who uses his love of music to help educate and inspire. Perhaps more important than these laudable scientific achievements, LeDoux is also a true “gentleman” of science, showing that science need not be a contact sport. Here, I give a personal account on why Joseph LeDoux is one of my scientific role models.


Fig. 1. Disambiguation of cue associations depends on DG context ensembles. (A) Context-odor paired associate task. Mice (N = 8) were trained concurrently in two spatial contexts. in context 1, digging in peppermint-scented bedding was reinforced, whereas in context 2, digging in carvone-scented bedding was reinforced. (B) Performance improved over training days [shading represents 95% confidence interval (Ci)]. (C) Percent time spent digging and (D) discrimination score in probe test after training (box represents 95% Ci; dashed line represents chance performance). (E) left: AAv-RAM-hM4di was microinjected into the dG. Middle: Removal of doxycycline (dOX) permitted hM4di tagging of dG neuronal ensemble active during pre-exposure to one of the to-be-discriminated contexts. After training, mice (N = 8) were probed in a tagged context after either vehicle (veh) or C21 treatment. Right: hM4di-expressing (red) neurons in dG [blue = 4′,6-diamidino-2-phenylindole (dAPi)]. Scale bar, 500 μm. (F) Performance improved across training (shading represents 95% Ci). (G) Percent time spent digging and (H) discrimination score in probe test after training (box represents 95% Ci; dashed line represents chance performance). *P < 0.05 (group comparison); # P < 0.05 (comparison to zero).
Fig. 2. Activation of dentate gyrus context ensembles reinstates context-specific memory retrieval. (A) AAv-RAM-hM3dq was microinjected into dG. (B) left: Removal of dOX permitted hM3dq tagging of dG neuronal ensemble active during pre-exposure to one of the to-be-discriminated contexts. After training, memory was probed in a novel context or the non-tagged context after either veh or C21 treatment. Right: hM3dq-expressing (red) neurons in the dentate gyrus (blue = dAPi). Scale bar, 500 μm. (C) Performance improved across training (shading represents 95% Ci) (N = 8). (D) Percent time spent digging and (E) discrimination score in probe test in the novel context. (F) Percent time spent digging and (G) discrimination score in probe test in the non-tagged context (box represents 95% Ci; dashed line represents chance performance). *P < 0.05 (group comparison); # P < 0.05 (comparison to zero).
Fig. 3. Experience-dependent plasticity of CA1 contextual representations. (A and B) AAv-RAM-hM3dq was microinjected into dG in thy1-GCaMP6f mice, and a miniature microscope was implanted above CA1. dOX removal permitted hM3dq tagging of dG neuronal ensembles active during context pre-exposure. After training, memory was probed in novel and non-tagged contexts after either veh or C21 treatment. (C) the dotted lines indicate a miniature microscope lens implanted above CA1 (blue = dAPi; green = GCaMP6f + cells). Scale bar, 500 μm. (D) example of CA1 imaging field across sessions. Randomly selected cells are colored by cell identity, and their corresponding denoised fluorescence traces are plotted below. (E) Performance improved across training (shading represents 95% Ci) (N = 6). (F and G) After training, memory was probed in novel (F) and non-tagged (G) contexts. in both contexts, C21 treatment increased digging time in the well which was reinforced in the tagged context during training. (H) CA1 population activity structure across training revealed by dimensionality reduction (trajectories plotted on a reduced set of dimensions using laplacian eigenmaps). (I) For each training day, the average cosine distance was computed between Pvs of trials within the same context (D within ) and between Pvs of trials in different contexts (D between ). Pv distance between contexts increased relative to distance within a context across training (blue = actual). Gray lines indicate distance ratios with randomized context ids (shuffle) (N = 6 mice). (J) latent-space distance between Pvs from different contexts increased with training. For each training day, Pvs were projected to a five-dimensional space using principal components analysis, and the distance between the five-dimensional distributions for each context was calculated (blue = actual). Gray lines and dots indicate chance level distances calculated by randomizing Pv context identity (shuffle) (N = 6 mice). a.u., arbitrary unit; n.s. not significant. *P < 0.05, **P < 0.01, and ***P < 0.001.
Fig. 4. Activation of DG context ensemble does not reinstate long-time scale population activity features of CA1 context-specific neural states. (A and B) For each context and training day, Pvs of average activity were calculated. Blue dots indicate the average similarity between a given day's Pv and the day 10 Pv from the same context. Gray dots signify the similarity between a given day's Pv and the day 10 Pv from the different contexts (n = 217 to 280 cells per session, N = 6 mice). (C and D) dG ensemble activation does not shift the population of active cells to a tagged context-like state in either (C) novel (n = 202 to 278 cells per session, N = 6 mice) or (d) nontagged (n = 162 to 251 cells per session, N = 6 mice) contexts. Pvs of average activity were calculated for all veh and C21 trials, and their correlation to training day 10 Pvs from the tagged (blue) and non-tagged (gray) contexts was calculated. (E and F) For each context and training day, a pairwise normalized correlation matrix (nCM) was computed (see Materials and Methods). dots indicate average similarity between a given day's nCM and the day 10 nCM from the same context (blue) or opposite context (gray) (n = 217 to 280 cells per session, N = 6 mice). (G and H) dG activation does not change the pattern of average pairwise functional connectivity in either (G) novel (n = 202 to 278 cells per session, N = 6 mice) or (h) non-tagged context (n = 162 to 251 cells per session, N = 6 mice). nCMs of average pairwise correlation were calculated for all veh and C21 trials, and their correlation to training day 10 nCMs from the tagged (blue) and non-tagged (gray) contexts was calculated. # P < 0.06, *P < 0.05, **P < 0.01, and ***P < 0.001.
Dentate gyrus ensembles gate context-dependent neural states and memory retrieval

August 2024

·

30 Reads

·

1 Citation

Science Advances

Memories of events are linked to the contexts in which they were encoded. This contextual linking ensures enhanced access to those memories that are most relevant to the context at hand, including specific associations that were previously learned in that context. This principle, referred to as encoding specificity, predicts that context-specific neural states should bias retrieval of particular associations over others, potentially allowing for the disambiguation of retrieval cues that may have multiple associations or meanings. Using a context-odor paired associate learning paradigm in mice, here, we show that chemogenetic manipulation of dentate gyrus ensembles corresponding to specific contexts reinstates context-specific neural states in downstream CA1 and biases retrieval toward context-specific associations.


Higher-order interactions between hippocampal CA1 neurons are disrupted in amnestic mice

July 2024

·

51 Reads

·

5 Citations

Nature Neuroscience

Across systems, higher-order interactions between components govern emergent dynamics. Here we tested whether contextual threat memory retrieval in mice relies on higher-order interactions between dorsal CA1 hippocampal neurons requiring learning-induced dendritic spine plasticity. We compared population-level Ca2⁺ transients as wild-type mice (with intact learning-induced spine plasticity and memory) and amnestic mice (TgCRND8 mice with high levels of amyloid-β and deficits in learning-induced spine plasticity and memory) were tested for memory. Using machine-learning classifiers with different capacities to use input data with complex interactions, our findings indicate complex neuronal interactions in the memory representation of wild-type, but not amnestic, mice. Moreover, a peptide that partially restored learning-induced spine plasticity also restored the statistical complexity of the memory representation and memory behavior in Tg mice. These findings provide a previously missing bridge between levels of analysis in memory research, linking receptors, spines, higher-order neuronal dynamics and behavior.


Comparing behaviours induced by natural memory retrieval and optogenetic reactivation of an engram ensemble in mice

June 2024

·

37 Reads

·

2 Citations

Memories are thought to be stored within sparse collections of neurons known as engram ensembles. Neurons active during a training episode are allocated to an engram ensemble (‘engram neurons’). Memory retrieval is initiated by external sensory or internal cues present at the time of training reactivating engram neurons. Interestingly, optogenetic reactivation of engram ensemble neurons alone in the absence of external sensory cues is sufficient to induce behaviour consistent with memory retrieval in mice. However, there may exist differences between the behaviours induced by natural retrieval cues or artificial engram reactivation. Here, we compared two defensive behaviours (freezing and the syllable structure of ultrasonic vocalizations, USVs) induced by sensory cues present at training (natural memory retrieval) and optogenetic engram ensemble reactivation (artificial memory retrieval) in a threat conditioning paradigm in the same mice. During natural memory recall, we observed a strong positive correlation between freezing levels and distinct USV syllable features (characterized by an unsupervised algorithm, MUPET (Mouse Ultrasonic Profile ExTraction)). Moreover, we observed strikingly similar behavioural profiles in terms of freezing and USV characteristics between natural memory recall and artificial memory recall in the absence of sensory retrieval cues. Although our analysis focused on two behavioural measures of threat memory (freezing and USV characteristics), these results underscore the similarities between threat memory recall triggered naturally and through optogenetic reactivation of engram ensembles. This article is part of a discussion meeting issue ‘Long-term potentiation: 50 years on’.



Development of a double trauma model for PTSD in mice
a Experimental schedule for memory retention tests against a single 2 s foot shock. The time to enter the dark compartment was recorded. b The latency in Context A (CtxA) shows inverted U-shape in response to a single foot shock (0, 0.5, 1, 2, or 3 mA) (Kruskal-Wallis test, H4 = 40.82, P < 0.0001; *Mann-Whitney U planned comparisons: 2 mA versus 3 mA, U = 30, P = 0.019). c The latency in CtxB, which was different in wall shape and smell compared to CtxA, increased depends on shock intensity (Kruskal-Wallis test, H4 = 33.31, P < 0.0001; Mann-Whitney U planned comparisons (corrected α = 0.016): 0.5 mA versus 3 mA, U = 6, P < 0.0001; 1 mA versus 3 mA, U = 30, P = 0.019; 2 mA versus 3 mA, U = 50, P = 0.26). d The latency was tested as CtxA1 after the 3-mA foot shock in CtxA, and CtxA2 and CtxB after the reminder foot shocks (RS; weak 1 mA or strong 3 mA) in a different context. e A strong reminder shock of 3 mA increased the latency in CtxA (Repeated measures ANOVA, Reminder × Day interaction: F1,11 = 7.46, P = 0.020). f A strong reminder shock of 3 mA modestly increased the latency in CtxB compared to a weak reminder shock of 1 mA (Unpaired t-test, t11 = 1.60, P = 0.13). g Anxiety-like behavior in the open field test (OFT) was assessed one day after the reminder shock s. h A strong reminder shock of 3 mA decreased center time (%) compared with a weak reminder shock of 1 mA (Kruskal-Wallis test, H2 = 13, P = 0.0016; Mann-Whitney U planned comparisons: 1 mA versus 3 mA, U = 3, P = 0.0009. i Entropy did not differ across reminder shock groups (Kruskal-Wallis test, H2 = 4.4, P = 0.11). j Experimental schedule for memory extinction from one day after the reminder shock (1 mA or 3 mA). A reminder shock of low intensity (1 mA) was administrated after extinction to assess memory reinstatement (on day 9). k The strong reminder shock group showed slower extinction (context A; day 3–7) and high level of memory reinstatement (context A; day 9) (Mann-Whitney U planned comparisons for 1 mA versus 3 mA (corrected α = 0.0083): Day 3, U = 18, P = 0.0011; Day 4, U = 2, P < 0.0001; Day 5, U = 5, P < 0.0001; Day 6, U = 17, P = 0.0009; Day 7, U = 30, P = 0.015; Day 9, U = 9.5, P < 0.0001). Data are represented as mean ± S.E.M. Data points represent individual mice (b, c: 3 mA, n = 14 mice, other groups, n = 10 mice; e, f: 3–1 mA, n = 6 mice, 3–3 mA, n = 7 mice; h, i; n = 8 mice for each group; k: n = 12 mice for each group). Statistical significance is indicated by asterisks.
PTSD-like symptoms persist at remote delays
a A reminder shock was administered 1 day after the training in Context A (CtxA), and latency was tested 4 weeks later. b No difference in CtxA latency 4 weeks after a reminder shock (Mann-Whitney U-test, U = 19, P = 0.53). c High intensity (3 mA) reminder shock group mice showed increased latency in CtxB (Unpaired t-test, t12 = 2.85, P = 0.015). d Center time (%) in the open field test was decreased in both of shock groups than no shock group (Kruskal-Wallis test, H2 = 18, P = 0.00013; Mann-Whitney U planned comparisons (corrected α = 0.016): Naïve versus 1 mA, U = 0, P = 0.0001; Naïve versus 3 mA, U = 0, P = 0.0001; 1 mA versus 3 mA, U = 12, P = 0.11). e The groups that received foot shocks exhibited decreased entropy compared with no shock group (Kruskal-Wallis test, H2 = 16, P = 0.00032; Mann-Whitney U planned comparisons (corrected α = 0.016): Naïve versus 1 mA, U = 3, P = 0.0007; Naïve versus 3 mA, U = 0, P = 0.0001; 1 mA versus 3 mA, U = 15, P = 0.26). f Experimental schedule for memory extinction 4 weeks after the reminder shock (1 mA or 3 mA). A reminder shock of low intensity (1 mA) was administrated after extinction to assess memory reinstatement in CtxA test on day 36. g High intensity reminder shock group showed slower extinction (context A; day 30–34) and high level of memory enhancement (context A; day 36) (Mann-Whitney U planned comparisons for 1 mA versus 3 mA (corrected α = 0.0083): Day 30, U = 39, P = 0.060; Day 31, U = 3, P < 0.0001; Day 32, U = 2, P < 0.0001; Day 33, U = 25.5, P = 0.0058; Day 34, U = 28, P = 0.0096; Day 36, U = 11, P = 0.0001). h A reminder shock and extinction training were administrated 4 weeks after the training in CtxA, and memory reinstatement was tested in CtxA on day 36 following a low intensity (1 mA) reminder shock. i, j High intensity reminder shock group showed increased latency in CtxA and CtxB (Mann-Whitney U-tests: i, U = 9, P = 0.014; j, U = 7, P < 0.0068). k Center time (%) in the open field test was decreased in both of shock groups compared with no shock group (Kruskal-Wallis test, H2 = 19, P < 0.0001; Mann-Whitney U planned comparisons (corrected α = 0.016): Naïve versus 1 mA, U = 0, P < 0.0001; Naïve versus 3 mA, U = 0, P < 0.0001; 1 mA versus 3 mA, U = 32, P = 0.99). l The groups that received foot shocks decreased entropy compared to no shock group (Kruskal-Wallis test, H2 = 16, P = 0.00040; Mann-Whitney U planned comparisons (corrected α = 0.016): Naïve versus 1 mA, U = 0, P = 0.0002; Naïve versus 3 mA, U = 0, P = 0.0002; 1 mA versus 3 mA, U = 17, P = 0.94). m Experimental schedule for memory extinction 1 day after the reminder shock (1 mA or 3 mA). Reminder shock of low intensity (1 mA) was administrated after extinction to assess memory reinstatement in CtxA test on day 36. n High intensity reminder shock group showed longer latency and memory enhancement compared to lower intensity group (Mann-Whitney U planned comparisons for 1 mA versus 3 mA (corrected α = 0.0083): Day 30, U = 10, P = 0.0001; Day 31, U = 16, P = 0.0006; Day 32, U = 17, P = 0.0009; Day 33, U = 46, P = 0.14; Day 34, U = 54, P = 0.32; Day 36, U = 6, P < 0.0001). Data are represented as mean ± S.E.M. Data points represent individual mice (b, c: n = 7 mice for each group; d, e: 0 mA, n = 10 mice, other groups, n = 7 mice for each group; g: n = 10 mice for each group; i, j: n = 8 mice for each group; k: n = 10 mice, other groups, n = 8 mice for each group; l: 0 mA, n = 10 mice, other groups, n = 6 mice for each group; n: n = 12 mice for each group). Statistical significances are indicated by asterisks.
Exercise attenuates PTSD-like phenotypes
a Experimental schedule of CtxA memory extinction and reinstatement to test the effect of exercise after the reminder shock. CtxA memory was tested on five consecutive days from day 30, and reinstatement was tested 1 day after a reminder shock. b Post-training exercise weakened CtxA memory and memory reinstatement following a 1 mA reminder shock (Mann-Whitney U planned comparisons for Sedentary versus Running (corrected α = 0.0083): Day 30, U = 11, P = 0.028; Day 31, U = 11, P = 0.026; Day 32, U = 8, P = 0.010; Day 33, U = 4, P = 0.0019; Day 34, U = 16.5, P = 0.11; Day 36, U = 5, P = 0.0030). c Generalization in CtxB or anxiety-like behavior in open field test (OFT) were assessed in CtxB test 4 weeks after the reminder shock. d Running suppressed the latency in CtxB (Mann-Whitney U-test, U = 0, P = 0.0022). e Excercise modestly increased the center time (%) in the OFT (Mann-Whitney U-test, U = 14, P = 0.053). f Exercise increased entropy in the OFT (Unpaired t-test, t14 = 3.34, P = 0.0048). g Experimental schedule of CtxA memory extinction and reinstatement to assess the effect of exercise before the reminder shock. h The running group showed rapid extinction and suppressed reinstatement (Mann-Whitney U planned comparisons for Sedentary versus Running (corrected α = 0.0083): Day 30, U = 31, P = 0.95; Day 31, U = 3, P = 0.0011; Day 32, U = 2, P = 0.0006; Day 33, U = 10, P = 0.020; Day 34, U = 4, P = 0.0017; Day 36, U = 0, P = 0.0002). i Generalization or anxiety-like behavior was tested one day after the reminder shock, which was administered 4 weeks after the CtxA training. j Exercise suppressed the latency in CtxB (Mann-Whitney U-test, U = 0, P = 0.0002). k Exercise increased center time (%) in the OFT (Mann-Whitney U-test, U = 21, P = 0.0053). l Exercise did not affect entropy in the OFT (Unpaired t-test, t20 = 1.76, P = 0.094). Data are represented as mean ± S.E.M. Data points represent individual mice (b: n = 8 mice for each group; d: n = 6 mice for each group; e, f: n = 8 mice for each group; h, j: n = 8 mice for each group; k, l: Sed, n = 10 mice, Run, n = 12 mice). Statistical significance is indicated by asterisks.
Hyper-integration of adult-generated granule cells weakens PTSD-like behaviors
a Mice expressed ChR2-eYFP (ChR2⁺) or control (ChR2⁻) in nestin positive cells received photo-stimulation (3 × 1 min/day at 10 Hz) for 14 days. Neuronal re-organization was induced by optogenetic stimulation in ChR2⁺ mice after foot shock administration. ChR2⁻ control mice also receive foot shocks and photo-stimulation. b, c ChR2⁺ mice showed decreased CtxA and CtxB latency 4 weeks after the foot shocks compared to ChR2⁻ control (Mann-Whitney U-tests: b, U = 9, P = 0.0002; c, U = 19, P = 0.010). d No difference in center time (%) of the open field test between ChR2⁺ and ChR2⁻ mice (Mann-Whitney U-test, U = 40, P = 0.31). e ChR2⁺ mice tended to increase spatial entropy than ChR2⁻ mice (Mann-Whitney U-test, U = 23, P = 0.024). f Sema5A deletion was induced by tamoxifen injection in Sema5A-/- mice. Neuronal re-organization was induced by tamoxifen injection after reminder foot shock in Sema5A-/- mice. The latency in CtxA and CtxB was decreased in the Sema5A deletion group (Mann-Whitney U-tests: (g) U = 9, P = 0.016; (h) U = 11, P = 0.029. (i) No difference in center time (%) in the open field test between control and Sema5A-/- mice (Mann-Whitney U-test, U = 14, P = 0.071). j Sema5A deletion increased spatial entropy in the open-field test (Unpaired t-test, t14 = 2.57, P = 0.022). Data are represented as mean ± S.E.M. Data points represent individual mice (b, c, d, e; ChR2⁻ mice n = 11, ChR2⁺ mice n = 10, g, h, i, j; Sema5A+/+ mice, n = 7, Sema5A-/- mice, n = 9). Statistical significance indicated by asterisks.
Post-training increases in hippocampal neurogenesis weakens conditioned place preference for cocaine
a One side of the conditioned place preference (CPP) box was conditioned with saline, and the other side was conditioned with cocaine (7.5 mg/kg) for 3 days training. CPP score was measured 4 weeks after the final training. b Exercise suppressed the CPP score (Unpaired t-test, t39 = 2.58, P = 0.014). c Nestin⁺ cells were stimulated optogenetically after CPP training to promote re-organization of hippocampal neuronal circuits in Nestin-CreERT2 ChR2 mice. d The ChR2⁺ group showed a decreased CPP score (Mann-Whitney U-test, U = 57, P = 0.0037). e Neuronal re-organization was induced by tamoxifen injection in Nestin-CreERT2 Sema5A-/- mice after CPP training. f Sema5A-/- group decreased CPP score (Unpaired t-test, t16 = 2.58, P = 0.020). Data are represented as mean ± S.E.M. Data points represent individual mice (b; Sed, n = 20, Run, n = 21, d; ChR2⁻ mice n = 16, ChR2⁺ mice n = 17, f; Sema5A+/+ mice, n = 11, Sema5A-/- mice, n = 7). Statistical significance is indicated by asterisks.
Neurogenesis-dependent remodeling of hippocampal circuits reduces PTSD-like behaviors in adult mice

May 2024

·

213 Reads

·

2 Citations

Molecular Psychiatry

Post-traumatic stress disorder (PTSD) is a hypermnesic condition that develops in a subset of individuals following exposure to severe trauma. PTSD symptoms are debilitating, and include increased anxiety, abnormal threat generalization, and impaired extinction. In developing treatment strategies for PTSD, preclinical studies in rodents have largely focused on interventions that target post-encoding memory processes such as reconsolidation and extinction. Instead, here we focus on forgetting, another post-encoding process that regulates memory expression. Using a double trauma murine model for PTSD, we asked whether promoting neurogenesis-mediated forgetting can weaken trauma memories and associated PTSD-relevant behavioral phenotypes. In the double trauma paradigm, consecutive aversive experiences lead to a constellation of behavioral phenotypes associated with PTSD including increases in anxiety-like behavior, abnormal threat generalization, and deficient extinction. We found that post-training interventions that elevate hippocampal neurogenesis weakened the original trauma memory and decreased these PTSD-relevant phenotypes. These effects were observed using multiple methods to manipulate hippocampal neurogenesis, including interventions restricted to neural progenitor cells that selectively promoted integration of adult-generated granule cells into hippocampal circuits. The same interventions also weakened cocaine place preference memories, suggesting that promoting hippocampal neurogenesis may represent a broadly useful approach in hypermnesic conditions such as PTSD and substance abuse disorders.


Citations (70)


... These engrams are thought to be a possible neural substrate for memory, and their formation depends on neural excitability levels at the moment of acquisition (Rashid et al., 2016;Ramsaran et al., 2023). As both glucocorticoid injections or acute stress have been shown to increase engram size and enhance fear generalisation (Lesuis et al., 2021;Lesuis et al., 2025), we set out to determine whether ELS results in memory changes due to alterations in putative engram recruitment. We collected brain tissue 90 min after conditioning ( Figure 2I) and carried out immunostaining for the immediate early gene c-Fos, a commonly used activity marker, in the lateral amygdala ( Figure 2J-L, Figure 2-figure supplement 2D and E). ...

Reference:

Early-life stress induces persistent astrocyte dysfunction associated with fear generalisation
Stress disrupts engram ensembles in lateral amygdala to generalize threat memory in mice
  • Citing Article
  • November 2024

Cell

... In support of this notion, evidence of a cellular mechanism whereby c-Fos levels inversely correlate with memory accuracy has been recently shown in hippocampal circuits (Asok et al., 2018). In addition, early-life experience strongly shapes episodic memory development, with stress delaying maturation of memory processes (Ramsaran et al., 2024). Early-life stress has even been shown to activate specific neuronal ensembles that contribute to stress hypersensitivity (Balouek et al., 2023). ...

A sensitive period for the development of episodic-like memory in mice

... In the hippocampus, the CA1 and CA3 regions strongly contribute to memory encoding and consolidation. 24,25 During SWS in particular, CA3 pyramidal neurons are spontaneously activated in synchronous bursts that trigger massive activation of CA1 pyramidal cells, promoting the strength of connections between clusters and ultimately completing the consolidation of memories. Moreover, hippocampal levels of the memory-associated transcription factors CREB and cAMP are elevated during REM sleep. ...

Higher-order interactions between hippocampal CA1 neurons are disrupted in amnestic mice

Nature Neuroscience

... This could resolve specific changes at engram synapses relating to different phases of memory encoding, such as recall, consolidation and storage, and relating to different brain regions 5,21,80,89 . Moreover, the combination of our workflow with both natural or artificial optogenetic memory recalls could enable the comparison of engram synapses between active and inactive memory circuits [90][91][92][93] . ...

Comparing behaviours induced by natural memory retrieval and optogenetic reactivation of an engram ensemble in mice

... A host of experimental evidence supports the hypothesis that synaptic plasticity is 9 essential for memory storage. However, some recent results indicate that also 10 non-synaptic plasticity such as the regulation of neuronal membrane properties 11 contributes to the creation of memory engrams [4][5][6][7][8]. In fact, there has been some 12 scepticism about the role of synaptic plasticity in memory formation [6,9,10]. ...

Neuronal Excitability in Memory Allocation: Mechanisms and Consequences
  • Citing Chapter
  • September 2020

... Since relative excitability has been shown to be a major determinant of neuronal allocation to a coding ensemble in the DG 88 , as well as in the CA1 region and amygdala 89,90 , cells that are most affected by adversity in the epigenetically and transcriptionally heterogenous population of MIA vDGCs may have increased excitability. However, expression of FOS and other IEGs was comparable in MIA and CON FOS+ cells (e.g., were not DEGs) suggesting no apparent difference in neuronal activity between the MIA and CON FOS+ DGC populations. ...

Excitability mediates allocation of pre-configured ensembles to a hippocampal engram supporting contextual conditioned threat in mice
  • Citing Article
  • March 2024

Neuron

... Recent technological advancements, including FLiCRE technology (but see also references for Cal-Light and FLARE [26][27][28][29], have paved the way for differentiating between and tagging cell populations with higher temporal resolution. This was unachievable with drug-and IEG-based engram tagging, considering that both the timing of drug delivery and that of IEG-derived protein production occur at a larger time scale than the acquisition window and that the expression of IEGs changes across brain regions, tasks, and moments 8,30,31 . ...

Examining memory linking and generalization using scFLARE2, a temporally precise neuronal activity tagging system

Cell Reports

... Running wheels were placed in the home cage 1-2 d after training. Mice voluntarily ran on the wheels as described previously [14,17,34]. Running wheels remained in the cage until the day before testing began (~28 d). ...

Neurogenesis-mediated circuit remodeling reduces engram reinstatement and promotes forgetting

... Fear learning and memory in the lateral amygdala have also been correlated with the allocation of a small number of neurons to a fear engram (Josselyn et al., 2015). These engrams are thought to be a possible neural substrate for memory, and their formation depends on neural excitability levels at the moment of acquisition (Rashid et al., 2016;Ramsaran et al., 2023). As both glucocorticoid injections or acute stress have been shown to increase engram size and enhance fear generalisation (Lesuis et al., 2021;Lesuis et al., 2025), we set out to determine whether ELS results in memory changes due to alterations in putative engram recruitment. ...

A shift in the mechanisms controlling hippocampal engram formation during brain maturation
  • Citing Article
  • May 2023

Science

... As a crucial brain structure involved in memory and knowledge transfer, the hippocampus has been extensively studied in order to understand its role in facilitating learning across various cognitive domains [6][7][8][9][10]. This process of knowledge transfer shares similarities with the concept of transfer learning in RL algorithms, and successor representation (SR) has emerged as an explainable algorithmic theory that may provide insights into the function of the hippocampus in transfer learning [7,[11][12][13]. ...

Emergence of a predictive model in the hippocampus
  • Citing Article
  • April 2023

Neuron